What is the optimum design for a hospital’s emergency department? How do you make the most of limited resources while facing huge demand? The Health Economics Unit was asked to help hospital trusts better understand how their current design works and what the impact of potential changes will be.

Using data to understand patient flow

Staffing shortages, increased patient volumes, and limited resources are placing extreme pressure on emergency departments nationally.

Making operational changes within an emergency care department can be costly. Simulations are an inexpensive and proactive way of testing the impact of operational changes before they are introduced. Identifying the root causes could not only help to improve quality and safety for patients undergoing emergency care but may also have wider cost-saving benefits.

The Health Economics Unit (HEU) was, therefore, tasked with developing an Urgent and Emergency Care (UEC) or continuous flow model to assess the impact of any operational changes before they are introduced.

Complex analytics for a complex system

The HEU team adapted a pre-existing UEC flow model to increase its complexity and make it more representative of the real-world, developing a discrete-event simulation. This type of modelling mimics what happens to patients as they pass through accident and emergency (A&E) departments e.g. triage, assessment, diagnostics, treatment, discharge and admission. Other urgent care departments, such as same day emergency care (SDEC) and urgent treatment centres (UTCs) were also modelled.

By using a discrete-event simulation, analysts can assess the impact resource capacity limitations may have e.g. a lack of space, prioritisation of clinical care based on urgency of care need (acuity), etc.

The existing UEC flow model was reconfigured to mimic a current hospital pathway. The complex code was developed by multiple analysts in Python, with version control using GitHub. Input from operational stakeholders was also included. The model considered different acuity of patients passing through different areas of the UEC system, e.g. resus versus majors versus minors versus SDEC. It handled varying demand by the time of the day and with variable processes, such as assessment, diagnostics, treatment, admission/discharge all having variable durations, depending upon acuity, with uncertainty explicitly modelled using a Monte-Carlo simulation.

Scenario modelling/sensitivity analyses were performed to compare results against a baseline model. A technique called process mapping was used to validate the simulated pathway.

Helping decision makers plan impactful changes

The UEC flow model has significant will help decision-makers understand the impact of operational changes on patient flow through emergency care departments, allowing them to better plan improvements and best use of budgets. The model (Figure 1) can be further developed to reflect system pathways, including those that may be implemented in the future. Additional components that could be included are:

• Frailty assessment units (FAU) and other key urgent areas
• Wider population information e.g., the type of presentation e.g. self-presentation versus
  GP/111-referrals, chief complaint, etc.
• Downstream effects e.g., bed occupancy, mental health coordination etc.
• Other wider determinants of health, e.g., the impact of age or multiple deprivation on types of presentations to the emergency department.

Talk to us about how we can help with flow modelling and other Population Health Management techniques. Contact us here.